Path planning of a variable scale tetrahedral rolling robot with autonomous mobility

In the field of deep space exploration, land inspection and detection robots are a crucial tool for detecting the terrestrial environment on the surface of rocky planets. With various complex terrain environments on the surface of different stars, traditional robots, such as wheeled robots, crawler robots, and legged robots, encounter the risk of overturning. The variable scale tetrahedral rolling robot with autonomous mobility is an innovative land inspection and detection robot that can achieve autonomous movement, continuous rolling motions and variable scale through its own spatial structure transformation. It has the advantages of high stability and flexibility. Path planning serves as a fundamental requirement for robots to achieve optimal movement. In contrast to traditional mobile robots, the trajectory formed by the tetrahedral rolling robot's movement is composed of triangular meshes. Currently, the path planning algorithms commonly used for mobile robots are not applicable to the path planning of tetrahedral rolling robot. To tackle the above problems and to meet the requirements of tetrahedral rolling robot's path planning, a genetic algorithm (GA)-based path planning approach is proposed. According to the mechanical structure of the robot, a form of rolling motion is proposed. Based on this motion principle, the trajectory as a triangular mesh is verified. Furthermore, the pose coordinates of the robot are established in a triangular grid map, while the transformation relationship between the triangular grid map and the Cartesian coordinate system is presented. According to the triangular mesh map, a path planning method based on GA is developed. Finally, the proposed method is verified experimentally by using MATLAB software. The experimental results illustrate that, the proposed method can realize the optimal path planning. The path planning method implemented in this paper lays a crucial foundation for the motion control of the robot. It provides a theoretical basis for robotic land exploration of complex environments in deep space exploration.